The phenazine pyocyanin is a terminal signalling factor in the quorum sensing network of Pseudomonas aeruginosa

The phenazine pyocyanin is a terminal signalling

factor in the quorum sensing network of Pseudomonas

aeruginosa

Lars E. P. Dietrich,1_{Alexa Price-Whelan,}2 Ashley Petersen,4_{Marvin Whiteley}4_and Dianne K. Newman1,2,3_*

1_{Divisions of Geological and Planetary Sciences and} 2_{Biology, and}3_{Howard Hughes Medical Institute,} California Institute of Technology, Pasadena, CA 91125, USA.

4_{Department of Microbiology and Immunology, The} University of Oklahoma Health Sciences Center, Oklahoma City, OK 73104, USA.

Summary

Certain members of the fluorescent pseudomonads produce and secrete phenazines. These heterocyclic, redox-active compounds are toxic to competing organisms, and the cause of these antibiotic effects has been the focus of intense research efforts. It is largely unknown, however, how pseudomonads themselves respond to – and survive in the presence of – these compounds. Using Pseudomonas aerugi-nosa DNA microarrays and quantitative RT-PCR, we demonstrate that the phenazine pyocyanin elicits the upregulation of genes/operons that function in trans-port [such as the resistance-nodulation-cell division (RND) efflux pump MexGHI-OpmD] and possibly in redox control (such as PA2274, a putative flavin-dependant monooxygenase), and downregulates genes involved in ferric iron acquisition. Strikingly, mexGHI-opmD and PA2274 were previously shown to be regulated by the PA14 quorum sensing network that controls the production of virulence factors (including phenazines). Through mutational analysis, we show that pyocyanin is the physiological signal for the upregulation of these quorum sensing-controlled genes during stationary phase and that the response is mediated by the transcription factor SoxR. Our results implicate phenazines as signalling molecules in both P. aeruginosa PA14 and PAO1.

Introduction

Bacteria communicate with each other through secreted signalling factors. These are small, diffusible molecules that are specifically released and then recognized by adjacent cells, where they trigger a response in gene expression. This ensures the behavioural synchronization of cell populations (Lazdunski et al., 2004). Because these events require high cell densities (presumably to accumulate sufficient signal), this mode of bacterial com-munication has been termed quorum sensing (QS). The best-characterized QS system is the Lux system of Vibrio fischeri, a luminous Gram-negative bacterium. The V. fischeri QS mechanism consists of a synthase (LuxI) that produces the signal, an acyl homoserine lactone (AHL) (Eberhard et al., 1981), and a transcriptional acti-vator (LuxR) that recognizes the signal, leading to the downstream activation of genes in the lux operon (Enge-brecht et al., 1983). Homologues of these genes have been identified in a number of Gram-negative bacteria, demonstrating the generality of QS signalling (Fuqua et al., 1994).

The Gram-negative bacterium Pseudomonas aerugi-nosa contains two such systems, Las and Rhl. LasI and RhlI synthesize the AHLs 3-oxo-dodecanoyl-homoserine lactone (3O-C12-HSL) and butanoylhomoserine lactone (C4-HSL) respectively, which in turn induce gene expres-sion by binding to their respective transcriptional activa-tors, LasR and RhlR (Passador et al., 1993; Pearson et al., 1995). Recently, a third signal has been found to participate in the P. aeruginosa QS network, the quinolone 2-heptyl-3-hydroxyl-4-quinolone that has been named the Pseudomonas quinolone signal (PQS) (Pesci et al., 1999). The production of PQS is under the regime of the QS network (McGrath et al., 2004). Its structure is very similar to the so-called Pyo compounds, which had been identified as antibiotics in 1945 (Hays et al., 1945) and shown to belong to the family of 4-quinolones in 1952 (Wells, 1952). The realization that PQS itself functions as a QS signal was surprising, because it is not an AHL and the genes involved in its synthesis are not LuxI homologues. However, apart from these differences, PQS fits all the functional criteria of a QS signal: (i) cell density-dependent accumulation (Lepine et al., 2003) of (ii) a Accepted 26 June, 2006. *For correspondence. E-mail dkn@gps.

caltech.edu; Tel. (+1) 626 395 6790; Fax (+1) 626 683 0621.

First published online 25 July 2006

small diffusible molecule that (iii) is recognized by adja-cent cells, (iv) in which it triggers a specific transcriptional response (Camilli and Bassler, 2006). This underscores the importance of defining QS systems on a functional level rather than on the basis of homology to canonical QS molecules. The three QS signal/response systems in P. aeruginosa are arranged in a temporal cascade, with AHLs and PQS being released in early and late exponen-tial phase respectively (Lepine et al., 2003). Until now, PQS has been thought to be the terminal signal in the P. aeruginosa QS cascade.

MvfR positively regulates the production of a number of virulence factors as well as expression of PA2274, a puta-tive monooxygenase, and the mexGHI-opmD operon that encodes proton-driven efflux pumps of the resistance-nodulation-cell division (RND) transporter superfamily (Deziel et al., 2005). It is thought that this response is mediated through PqsE and the signal PQS (Deziel et al., 2005). Among the virulence factors that are produced in response to PQS signalling are phenazines (Deziel et al., 2004). These are heterocyclic, redox-active compounds that exert toxic effects on other prokaryotes and eukary-otes (Mazzola et al., 1992; Mahajan-Miklos et al., 1999). Like quinolones, phenazines are excreted from cells at specific points following exponential growth (Diggle et al., 2003). The antibiotic properties of phenazines have been the subject of intense research efforts, but little is known about their effect on the producing organisms (Hassett et al., 1992). How, for example, do pseudomonads respond to these toxic compounds, abundantly secreted by cells during stationary phase? Moreover, why are these compounds secreted at this late growth stage? Previously, we hypothesized that phenazines could be serving impor-tant physiological functions under these conditions, sepa-rate from being antibiotics, such as facilitating extracellular electron transfer by being cycled in and out of cells (Her-nandez and Newman, 2001; Price-Whelan et al., 2006).

To gain insight into how phenazines might be function-ally integrated into P. aeruginosa physiology, we analysed their effect on gene expression in the opportunistic patho-gen P. aeruginosa (strains PA14 and PAO1). Here we report that the phenazine pyocyanin (PYO) is a terminal signalling molecule in the P. aeruginosa QS network, and

suggest that phenazines may, in this capacity, control their own cycling.

Results

PA14 and PA01 excrete phenazines in early stationary phase

Pseudomonas aeruginosa releases at least four phena-zines (Mavrodi et al., 2001). These are 1-hydroxyphena-zine (1-OH-PHZ), phena1-hydroxyphena-zine-1-carboxamide (PCN) and PYO, which are all derived from the common precursor, phenazine-1-carboxylic acid (PCA) (Budzikiewicz, 1993; Mavrodi et al., 2001; Price-Whelan et al., 2006) (Fig. 1). To characterize phenazine release from P. aeruginosa strains PA14 and PAO1, we grew them in minimal medium, removed culture aliquots at different time points, and measured cell densities and the phenazine concen-tration in cell-free supernatants by high-performance liquid chromatography (HPLC). Under these conditions, phenazine release by PAO1 was below our detection limit, but PA14 released phenazines in early stationary phase (Fig. 2A). PYO, which is responsible for the blue colour of stationary phase cultures, and PCA were the predominant phenazines in cell-free supernatant by 10 h of growth, accumulating to approximately 30mM. We were unable to unequivocally detect 1-OH-PHZ and PCN under these conditions.

When grown in Luria–Bertani (LB), we observed PYO to be released by both strains, at the same stage in cell growth, although the concentration of PYO was~10 times greater for PA14 (~60 mM) than PAO1 (Fig. 2B).

Phenazines as signals: exogenous PYO elicits a specific genetic response in PA14

To gain insight into whether P. aeruginosa mounts a physi-ological response to phenazines, we investigated the effect of phenazines at the level of gene expression. Spe-cifically, we used Affymetrix P. aeruginosa GeneChips to identify transcriptional changes in PA14 in response to exogenous phenazines. We restricted our focus to PYO, one of the predominant phenazines in the supernatant of

Fig. 1. Phenazine biosynthesis in

P. aeruginosa. Genes encoding proteins

involved in phenazine production are listed above the arrows.

Chorismate phzA1-G1 phzA2-G2 phzH phzM/phzS phzS Phenazine-1-carboxylic acid (PCA) Phenazine-1-carboxamide (PCN) 1-hydroxyphenazine (1-OH-PHZ) Pyocyanin (PYO) © 2006 The Authors

stationary phase cultures. Purified PYO was added to PA14 cultures in early exponential phase (OD600= 0.4), well before the release of endogenous phenazines (Fig. 3A). Following the addition of PYO, the culture was incubated for 1 h, and total RNA was extracted, labelled, and hybridized to Affymetrix GeneChips arrayed with P. aeruginosa DNA representing the total genome. Pair-wise comparison of gene expression in the absence or presence of PYO was carried out using Affymetrix Microarray Suite software. Table 1 lists all genes that were up- or downregulated more than twofold in two inde-pendent experiments.

A number of features are striking: (i) PYO affects expres-sion of only a limited number of genes. Twenty-two genes are upregulated, eight of which encode putative transport-ers [i.e. the RND transporttransport-ers mexGHI-opmD (PA4205-4208), PA3923-3922 (mexEF homologues)-opmE and the putative Major Facilitator Superfamily (MFS) transporter PA3718]. Twenty-nine genes are downregulated, of which 16 are annotated as hypothetical proteins. Seven of the remaining 13 genes are implicated in ferric iron transport. (ii) The downregulated gene cluster PA5498-5500 encodes the transcriptional regulator np20 and proteins with similarity to zinc transporters. Interestingly, mutation of np20 was reported to abrogate PQS production, presum-ably due to a more than 10-fold decrease in pqsH expres-sion (Gallagher et al., 2002). This suggests that PYO might inhibit PQS production. (iii) The most upregulated genes, PA2274 (a putative monooxygenase) and the mexGHI-opmD operon, had previously been found to be downregu-lated in an mvfR mutant in PA14 (Deziel et al., 2005). It is thought that MvfR controls gene expression by regulating the production of PQS and PqsE (Deziel et al., 2005). Among the genes controlled by MvfR are phzM, phzS and the operon phzA1-G (Deziel et al., 2005). The requirement of PQS and PqsE has been well established for expression of phzA1-G1 (Diggle et al., 2003; Deziel et al., 2005). Based on these findings it could be assumed that phzA1-G1 and mexGHI-opmD are part of the same, PQS/ PqsE-responsive, regulon. However, it had also been noticed that while the phz operon was already expressed in (late) exponential phase, the expression of mexGHI-opmD occurred only in stationary phase of both PAO1 (Whiteley et al., 1999) and PA14 (Deziel et al., 2005). Our results may explain this finding: PQS signals phenazine production, which in turn upregulates PA2274 and mexGHI-opmD. This scenario predicts a fast gene expression response of mexGHI-opmD and PA2274 to PYO. To evaluate this pos-sibility, we added PYO to an exponential phase culture of PA14 and followed mexG expression over 1 h by quantita-tive reverse transcription PCR (Q-RT-PCR) (Fig. 3B). In contrast to the significant lag between expression of phz1 expression (which is controlled by PQS) and of mexGHI-opmD (Whiteley et al., 1999; Deziel et al., 2005), addition of 10 40 30 20 0 15

phenazine concentr

a

tion (µM)

OD (600 nm)

time [h]

OD (600 nm)

time [h]

p

y

ocy

anin concentr

a

tion (µM)

A

B

Phenazine release from PA14 in

MOPS minimal medium

PYO release from PA14 and PAO1 in LB

Fig. 2. PA14 and PAO1 release phenazines in early stationary phase.

A. PA14 grown in MOPS minimal medium releases phenazines in early stationary phase. PA14 was grown aerobically in a MOPS minimal medium with 20 mM succinate. Concentrations of phenazines were determined by HPLC analysis of filtrates from culture samples. Data are representative of at least three independent cultures.

B. PAO1 and PA14 grown in LB release PYO in early stationary phase. PA14 and PAO1 were grown aerobically in 50 ml of LB at 37°C. At indicated times, 1 ml was removed for measurement of optical density at 600 nm and PYO concentrations. Error bars represent standard deviations (n= 3).

+buffer

remove aliquot for RNA extraction after 5’, 10’, 20’, 30’, 40’, 50‘, 60’ +0.2mM PYO 0 0.5 1 1.5 2 2.5 2 4 6 8 10 time [h] OD (600 nm) 0 20 40 60 80 100 120 140 0 10 20 30 40 50 60 fold c hange in mexG e x pr ession PYO buffer control time [min] A B fo ld c h ange 0 20 40 60 80 100

clpX mexG PA2274 clpX mexG PA2274

Change in gene expr. from exp. to stat. phase Change in gene expr.

after addition of PYO to exp. phase

C PYO-dependent gene expression in PAO1

Fig. 3. A and B. PYO triggers a rapid response of mexG expression. PA14 was grown to OD600= 0.4 (exponential phase) in 100 ml of MOPS minimal medium supplemented with 20 mM succinate and split into two flasks. To one flask 0.2 mM PYO was added, to the other flask just buffer. After incubation for 5, 10, 20, 30, 40, 50 and 60 min, aliquots (3 ml) were removed for RNA extraction. mexG and clpX (control) expression were monitored by Q-RT-PCR. mexG expression levels were normalized to the constitutively expressed clpX. Changes in mexG expression are shown. The same results were obtained after normalizing mexG expression to the constitutively expressed gene recA (not shown).

C. PYO-dependent upregulation of gene expression in PAO1. PAO1 was grown aerobically in 60 ml of LB at 37°C to OD600= 0.3 (early exponential phase) and split into two flasks. To one flask 80mM PYO were added, to the other flask just buffer. After incubation for 1 h at 37°C (OD600= 0.5; exponential phase) 3 ml of aliquots was removed from both flasks for RNA extraction. Growth of the untreated culture was continued to an OD600= 2.0 (stationary phase), then another 3 ml of aliquot was removed for RNA extraction. After preparation of cDNA, Q-RT-PCR was performed to analyse expression of mexG, PA2274 and the constitutively expressed clpX and recA (controls). All expression levels were normalized to recA expression. Changes in gene expression after addition of PYO to exponential phase cultures are shown by the grey bars. Changes in gene expression from exponential to stationary phase are shown by the white bars. PYO addition experiments to exponential phase cultures; white bars represent the difference. The data represent the average of biological triplicates and error bars indicate the standard deviation.

PYO induced a rapid response of mexG expression (⬍ 5 min), suggesting that the corresponding operon is under the direct control of PYO. Similarly, addition of PYO to exponential phase cultures of PAO1 induced expression of mexG and PA2274 (Fig. 3C).

Because quinolones (such as PQS) are not made in early exponential phase (Diggle et al., 2003; Lepine et al., 2003) and our array did not reveal PYO-induced upregu-lation of upstream QS genes, our data suggested that PYO alone, rather than PQS, is a signal for mexGHI-opmD and PA2274 expression. To directly test this, we first ruled out the possibility that PQS production was induced by PYO by measuring the expression of pqsH, which encodes the last enzyme in the biosynthetic pathway for PQS (Gallagher et al., 2002), in the presence or absence of PYO in exponential phase. As expected, no significant difference in expression of mexG (one of the strongly upregulated genes from the array) was observed; in contrast pqsH expression significantly increased from exponential to stationary phase (Fig. 4A). Moreover, we still observed significant increases in mexG and PA2274 expression in a pqsH deletion mutant background when PYO was added to exponential phase cultures (Fig. 4B).

Physiological upregulation of genes in the PYO regulon in stationary phase is dependent on phenazine

biosynthesis

If PYO were the physiological signal for the expression of genes identified in the array (Table 1), we would expect that these genes would be similarly regulated during sta-tionary phase (when phenazines are released; Fig. 2) and that this response would require the production of phena-zines. To test this, we deleted both phz operons in PA14 to prevent phenazine production (see Experimental proce-dures, Figs 1 and 5A). We verified the absence of phena-zines by HPLC. The double mutant (Dphz1/2) did not show a growth defect (Fig. 5B), ruling out any significant pleio-tropic effects. We then used Q-RT-PCR to compare Dphz1/2 and wild-type cultures for their ability to regulate expression of PYO-controlled genes in stationary phase compared with exponential phase. For this we chose to follow the expression of genes that were strongly (⬎ 35¥) and weakly (⬍10 ¥) upregulated (mexGH and PA2275 respectively) in the array (Table 1). We grew PA14 WT and PA14 Dphz1/2 in rich medium, removed aliquots during exponential and stationary phase, and extracted total RNA for Q-RT-PCR analysis. mexG, mexH and PA2275 were upregulated in stationary phase (Fig. 5C), and their relative expression levels corresponded to their response to exog-enously added PYO (Table 1). Similar results were found for PAO1 (Fig. 3C). In contrast, absence of phenazine production (Dphz1/2) prevented the increased expression of mexG, mexH and PA2275 in stationary phase. These Table 1. Exogenous PYO elicits a specific gene response in PA14.

PA number Fold change Gene/annotation

Genes upregulated after addition of PYO to PA14 in exponential phase

PA4205 36.8 mexG

PA4206 43.7 mexH

PA4207 30.9 mexI

PA4208 20.0 opmD

PA2274 30.4 Putative monooxygenase

PA3521 3.4 opmE

PA3522 6.3 76% similar to mexF (PA4294) PA3523 6.2 60% similar to mexE (PA4293) PA3515 5.3 Hypothetical protein

PA3516 5.4 Probable lyase PA3517 5.5 Probable lyase PA3518 6.6 Hypothetical protein PA3519 5.9 Hypothetical protein PA3920 4.2 Probable P-type ATPase PA3718 3.9 Probable MFS transporter PA3520 3.4 Hypothetical protein

PA3721 2.9 Probable transcriptional regulator PA2275 2.8 Probable alcohol dehydrogenase PA3720 2.7 Hypothetical protein

PA4878 2.3 Probable transcriptional regulator PA3531 2.1 bfrB (bacterioferritin)

PA2847 2.1 Hypothetical protein

Genes downregulated after addition of PYO to PA14 in exponential phase

PA3600 -20.4 Hypothetical protein PA3601 -9.0 Hypothetical protein PA4063 -7.6 Hypothetical protein PA0781 -6.7 Putative haemin receptor PA2386 -6.7 pvdA (L-ornithine N5-oxygenase)

PA2398 -6.7 fpvA (ferripyoverdine receptor)

PA2845 -6.7 Hypothetical protein PA5532 -6.7 Hypothetical protein PA4834 -6.6 Hypothetical protein PA4835 -6.5 Hypothetical protein PA4836 -6.2 Hypothetical protein

PA4837 -9.7 Putative ferrichrome iron receptor PA4514 -5.8 piuA, probable receptor for iron

transporter PA1924 -5.1 Hypothetical protein PA4064 -4.3 Probable component of ABC

transporter PA4065 -3.1 Hypothetical protein PA4066 -3.0 Hypothetical protein PA1921 -3.1 Hypothetical protein PA1925 -3.0 Hypothetical protein

PA4133 -3.0 Cytochrome c oxidase subunit

PA5531 -3.0 tonB

PA4838 -2.7 Hypothetical protein PA4134 -2.6 Hypothetical protein PA4170 -2.6 Hypothetical protein

PA4515 -2.4 piuC, putative iron-uptake factor

PA1922 -2.4 Probable TonB-dependent receptor PA5498 -2.0 znuA (ABC-type Zn2+_transport) PA5499 -2.5 np20 (transcriptional regulator) PA5500 -2.0 znuC (zinc transport protein ZnuC)

Cells were grown to OD600= 0.4 (exponential phase) in MOPS minimal medium supplemented with 20 mM succinate and split into two flasks. PYO, purified from stationary phase cultures of PA14, was added to one flask to a final concentration of~0.2 mM and buffer was added to the other as a control. The cells were incubated for 1 more hour at 37°C before RNA was harvested. The effects on gene expres-sion were monitored by microarray analysis using Affymetrix GeneChips. Shown are genes that were up- or downregulated twofold or more in two independent experiments.

results were supported by similar experiments with a mutant defective in PhzM production [PhzM encodes an enzyme necessary for the biosynthesis of PYO (Mavrodi et al., 2001; 2006)] (Fig. S1). We conclude that phenazines are necessary for the physiological induction of down-stream genes in the P. aeruginosa QS network.

PYO activates the transcription factor SoxR

It was recently demonstrated that P. aeruginosa PAO1 cells upregulate mexGHI-opmD, PA2274 and PA3718 upon exposure to paraquat (Kobayashi and Tagawa, 2004; Palma et al., 2005). All three genes/operons are also upregulated in the presence of PYO, suggesting that PYO and paraquat (a.k.a. ‘methyl viologen’, a small, redox active compound structurally similar to phenazines) affect the same transcriptional regulator. The response to paraquat is mediated through PA2273, which is homolo-gous to soxR in Escherichia coli (Kobayashi and Tagawa, 2004). SoxR activates genes containing a specific upstream binding site (Sox box). In P. aeruginosa unequivocal Sox boxes are only found upstream of mexGHI-opmD, PA2274 and PA3718 (Kobayashi and Tagawa, 2004; Palma et al., 2005). To test if PYO-dependent upregulation of these genes is mediated by SoxR, we deleted PA2273 (soxR) and used Q-RT-PCR to determine changes in gene expression from exponential to stationary phase (Fig. 6). PYO-controlled genes lacking a Sox box, such as PA3920 and the oprN homologue opmE (PA3521) were not affected by the soxR deletion. In contrast, upregulation of mexG and PA2274 in stationary phase was lost in the absence of SoxR (Fig. 6). This differential effect on PYO-regulated genes not only estab-lishes PYO as an endogenous activator of the SoxR regulon, but also demonstrates that additional transcrip-tion factor(s) must be involved in the PYO response for regulation of genes lacking a Sox box. The nature of these factors remains to be elucidated.

SoxR activation by PYO does not rely on superoxide generation

Previous work with E. coli has established that SoxR binds DNA as a homodimer, each monomer containing a [2Fe-2S] cluster (Gaudu et al., 1997; Hidalgo et al., 1997). These clusters are kept in a reduced state, rendering SoxR inactive. Only upon their oxidation does SoxR get activated (Ding et al., 1996; Gaudu and Weiss, 1996). A much-discussed mechanism of SoxR activation is the generation of superoxide, which can directly oxidize the [2Fe-2S] clusters. However, the necessity to generate superoxide for SoxR activation in vivo has been questioned. Given that PYO is a redox active compound with a standard redox-potential of-34 mV at pH 7 (Price-0 10 20 30 40 recA mexG pqsH exp.->stat. PYO added 1.0 1.2 38.1 25.4 10.5 1.3 fold c h ange fo ld c h a n g e 0 10 20 30 clpX mexG PA2274

Addition of PYO does not affect pqsH expression

PYO-dependent gene expression does not require PQS production

A

B

Fig. 4. PYO regulates gene expression in a PQS-independent manner.

A. PYO addition to PA14 does not affect pqsH expression. PA14 was grown and treated as described for Fig. 3C. Q-RT-PCRs were performed to analyse expression of pqsH and the constitutively expressed clpX and recA (controls). All expression levels were normalized to clpX expression. Changes in gene expression from exponential to stationary phase (white bars) and in response to PYO addition in exponential phase (grey bars) are shown. The data represent the average of biological triplicates and error bars indicate the standard deviation.

B. PYO-dependent gene expression does not require PQS production. PA14DpqsH was grown aerobically in 60 ml of LB at 37°C to OD600= 0.3 (early exponential phase) and split into two flasks. To one flask 80mM PYO was added, to the other flask just buffer. After incubation for 40 min at 37°C (OD600= 0.5; exponential phase) 3 ml were removed from both flasks for RNA extraction. Q-RT-PCRs were performed to analyse expression of mexG, PA2274 and the constitutively expressed clpX and recA (controls). All expression levels were normalized to recA expression. Changes in gene expression in response to PYO-addition (grey bars) are shown. The data represent the average of biological triplicates.

0.01 0.1 1 10 0 1 2 3 4 5 6 7 RNA extraction Q-RT PCR WT Δphz1/2 200 100 0 204 181 15

mexG mexH PA2275

3 7 4 WT Δphz1/2 fold c hange OD (600nm) time [h] phzM A1 B1 phzC1 D1 phzE1 F1 G1 phzS 4713000 4721600 phzH 65000 69000

A

C

B

Fig. 5. Upregulation of the PYO regulon in stationary phase is dependent on phenazine production.

A. Genomic localization of phz genes in P. aeruginosa. White: the core operons phzA1-G1 and phzA2-G2, which are responsible for PCA synthesis. Grey: monocistronic genes encoding for the phenazine-decorating enzymes PhzM, PhzS and PhzH.

B and C. PA14 WT and PA14DphzA1-G1/phzA2-G2 (Dphz1/2) were grown aerobically in 50 ml of LB at 37°C. At OD600= 0.4 (exponential phase) or OD600= 1.4 (stationary phase), 4 ml were removed for RNA extraction. After preparation of cDNA, Q-RT-PCR was performed to analyse expression of mexG, mexH, PA2275 and the constitutively expressed clpX and recA (controls). All expression levels were normalized to recA expression. Changes in gene expression from exponential to stationary phase are shown. The data represent the average of biological triplicates and error bars indicate the standard deviation.

Fig. 6. Upregulation of a subset of the PYO-regulated genes is mediated by the transcription factor SoxR. PA14 and PA14 DsoxR were grown aerobically in LB to OD600= 0.4 (exponential phase) or OD600= 1.4 (stationary phase). RNA was extracted, cDNA was prepared, and Q-RT-PCR was performed to analyse expression of mexG, mexH, mexE, PA3920 and the constitutively expressed clpX and recA (controls). All expression levels were normalized to recA expression. Changes in gene expression from exponential to stationary phase are shown. The data represent the average of biological triplicates and error bars represent the standard deviation.

Change in gene expression from exponential to stationary phase

fo

ld c

hange

Whelan et al., 2006), we hypothesized that PYO might be able to activate SoxR [whose midpoint potential is -285 mV (Ding et al., 1996; Kobayashi and Tagawa, 2004)] in a superoxide-independent manner. To address this, we grew PA14 under strictly anaerobic conditions with nitrate as electron-acceptor. In exponential phase we added PYO that was dissolved in anoxic medium, incu-bated the culture for an additional hour, and determined expression of mexG and mexH (SoxR-regulated) and PA2275 (not SoxR-regulated) by Q-RT-PCR. As Fig. 7 shows, PYO was able to upregulate expression of mexG and mexH. In agreement with our results under aerobic conditions (Table 1, Fig. 5), upregulation of mexGH was still more than five times higher compared with changes in expression of PA2275. This strongly suggests that PYO can activate SoxR in a superoxide-independent manner, and implies that the SoxR regulon does not provide a defence against superoxide per se in P. aeruginosa. Con-sistent with this, we observe no growth defect of a soxR

deletion mutant in aerobic PA14 cultures (Fig. S2), con-firming the findings of Palma et al. (2005) in PAO1.

Discussion

In this study, we have shown that the secreted phenazine PYO, commonly thought of as a virulence factor, is a signalling molecule for P. aeruginosa strains PA14 and PAO1. PYO is both necessary (Fig. 5) and sufficient (Table 1, Fig. 3C) to upregulate the expression of a limited set of genes that we now designate the PYO stimulon. Within this stimulon are genes that appear to be involved in efflux and redox processes, as well as iron acquisition. MexGHI-opmD and PA2274, which are the most upregu-lated genes in the PYO stimulon, were previously found to be regulated by the QS signalling molecule PQS and PqsE (Deziel et al., 2005). PQS also controls expression of the phz operon (Gallagher et al., 2002; Deziel et al., 2004), which encodes the machinery responsible for phenazine synthesis (Mavrodi et al., 2001). Our results demonstrate that it is the production of phenazines (spe-cifically PYO) that links PQS to mexGHI-opmD and PA2274 expression. This establishes PYO as a terminal physiological signal in the QS circuitry (Fig. 8).

The subset of the PYO stimulon that is most strongly upregulated maps directly to a set of genes whose expression had previously been shown to be induced by the redox-active compound paraquat (i.e. methyl violo-gen) (Palma et al., 2005). Our results provide genetic evidence that the transcription factor SoxR is responsible for activating this subset. Although the mechanism of PYO activation of SoxR remains to be elucidated, we find that such activation occurs under strictly anaerobic conditions. SoxR nitrosylation by nitric oxide has been demonstrated to be a mechanism for its activation (Touati, 2000), but this may not explain our findings because PYO is known to inactivate nitric oxide (Warren et al., 1990). Consistent with this, nitric oxide metabolite concentrations appear to be lower in the sputum of cystic fibrosis patients relative to the endotracheal secretions of healthy individuals (Grase-mann et al., 1998); intriguingly, significant phenazine pro-duction is also known to occur in sputum (Palmer et al., 2005). Taken together, our data and these observations suggest that PYO can directly activate SoxR, and we are currently working to test this hypothesis.

Recognizing that PYO can activate SoxR under anaerobic conditions prompts us to re-evaluate the con-ventional wisdom that SoxR responds to superoxide stress. Historically, most studies of SoxR have been per-formed in E. coli. In this organism, SoxR has only one target gene, soxS. It is positioned next to soxR and encodes a transcription factor that turns on more than 40 genes, many of which are of crucial importance for E. coli’s oxidative stress response to superoxide

(Pom-Change in gene expression after

PYO addition to anaerobic cultures

fo

ld c

hange

Fig. 7. PYO activates the SoxR regulon under anaerobic conditions. PA14 was grown anaerobically in minimal medium supplemented with 20 mM succinate (carbon source) and 100 mM potassium nitrate (electron acceptor) at 37°C. At OD600= 0.4 (exponential phase) 0.2 mM PYO (in anaerobic minimal medium) were added. After 1 h incubation RNA was extracted, cDNA was prepared, and Q-RT-PCR was performed to analyse expression of

mexG, mexH, PA2275 and the constitutively expressed recA and clpX (control), to which expression levels were correlated. The

changes in gene expression from PYO-treated compared with buffer-treated cultures are shown. The data represent the average of biological triplicates and error bars indicate the standard deviation (except for PA2275 and recA; n= 2).

posiello et al., 2001). As a consequence, mutations in E. coli soxR mutants confer increased sensitivity to super-oxide generating compounds. Pseudomonas lacks a soxS homologue, however, and SoxR from P. putida and P. aeruginosa does not regulate genes involved in the oxidative stress response (Kobayashi and Tagawa, 2004; Palma et al., 2005; Park et al., 2006). Because deleting soxR does not significantly affect viability of P. aeruginosa PAO1 (Palma et al., 2005) and PA14 (Fig. S2), this argues for the existence of a SoxR-independent oxidative stress response. A recent observation by Demple and co-workers fits these findings (Park et al., 2006). After deleting soxR in P. putida KT2240 they found that this mutant remained as resistant to the superoxide-generating compound paraquat as the wild-type strain.

If SoxR does not induce an oxidative stress response in P. aeruginosa, then what is the function of its regulon? The six genes controlled by SoxR comprise the putative monooxygenase (PA2274) and two transporters, i.e. the RND efflux pump MexGHI-OpmD and the MFS trans-porter PA3718. Based on sequence homology, PA2774 groups with ActVA-Orf6 from Streptomyces coelicolor, a monooxygenase that binds to and oxygenates actinor-hodin, an aromatic polyketide similar in structure to phenazines (Sciara et al., 2003). Interestingly, PhzS, the enzyme that participates in the conversion of PCA to PYO, is a monooxygenase as well (Mavrodi et al., 2001). We therefore hypothesize that PA2274 might rec-ognize phenazines, allowing it to either (i) execute its putative enzymatic activity, (ii) act as a competitor against the monooxygenase PhzS, (iii) function as a chaperone, shielding the cellular environment from the toxic phenazines, or (iv) sense phenazines, thereby affecting gene expression. MexGHI-OpmD is one of 10

potential RND pumps in P. aeruginosa (Stover et al. 2000). It has recently been postulated that this efflux pump is required for the excretion of a PQS precursor (Aendekerk et al., 2002; 2005). In this respect it is striking that expression of mexGHI-opmD increases in response to phenazines (Table 3), which act down-stream of PQS. We hypothesize that MexGHI-OpmD is not only implicated in the release of a toxic PQS pre-cursor (Aendekerk et al., 2005), but potentially also in the efficient shuttling of phenazines (possibly in conjunc-tion with the putative MFS transporter PA3718).

Phenazines and quinolones for a long time have been recognized for their antimicrobial activities. The fact that the quinolone PQS and now the phenazine PYO can act as intercellular signals underscores the possibility that ‘secondary’ metabolites may in fact be of primary impor-tance in tuning the cell’s response to particular physiologi-cal states. Indeed, a layered control network, such as the QS cascade in P. aeruginosa, appears to be optimally designed to permit this for a population. For example, phenazine production initiates under conditions of oxidant limitation. Once phenazines are produced, they can react with oxidized species other than molecular oxygen, such as ferric iron oxides, thus facilitating iron acquisition by reducing it to ferrous iron (Cox, 1986; Hernandez et al., 2004). It is therefore not surprising that a phenazine con-trols the expression of genes that appear to be involved in iron acquisition and redox homeostasis, as well as genes that likely control their own processing (mexGHI-opmD). We predict that this is but one of many cases where secondary metabolites will prove to have functions that transcend their antibiotic activities and enable the co-ordinated response of microbial communities to changes in their environment.

OH O N H N H O O

LasI LasR RhlI

MvfR PqsA-G RhlR PqsH upregulation of upregulation of phzA1-G1 (phenazines) hcnABC (hydr. cyanide) chiC (chitinase) lecA,B (lectin) N H O O O 3-oxo-C12-HSL C4-HSL PQS _CH3 OH N N+ PYO

exponential phase late exponential phase early stationary phase

upregulation of

mexGHI-opmD (efflux)

PA2274 (monooxygenase)

SoxR

Fig. 8. Model of the QS network in P. aeruginosa PA14. The PA14 QS network comprises a cascade of three types of signalling molecules that function in a growth stage-dependent manner. In exponential phase LasI and LasR synthesize the AHLs 3O-C12-HSL and

butanoylhomoserine lactone (C4-HSL) respectively, which in turn induce gene expression by binding to their respective transcriptional activators, LasR and RhlR. This results in the production and release of PQS, which is required for the synthesis of phenazines. The phenazine PYO activates the transcription factor SoxR, thereby upregulating the expression of mexGHI-opmD and PA2274 (modified from Price-Whelan et al., 2006).

Experimental procedures Strains and plasmids

Bacterial strains and plasmids that were used or made in this study are listed in Table 2.

Growth conditions

All strains were grown aerobically at 37°C unless otherwise specified. LB medium was used for routine culturing. For certain experiments, P. aeruginosa was grown in morpho-linepropanesulphonic acid (MOPS)-buffered medium [50 mM MOPS (pH 7.2), 93 mM NH4Cl, 43 mM NaCl, 3.7 mM

KH2PO4, 1 mM MgSO4and 3.5mM FeSO4·7H2O] with 20 mM

succinate (Palmer et al., 2005).

Construction of mutants

We generated unmarked deletions of PA2273 (soxR) and of the two redundant phenazine biosynthetic operons phzA1-G1 and phzA2-G2 in PA14. To make the double mutant DphzA1-G1/DphzA2-G2, we first deleted phzA1-G1 and then phzA2-G2. Here we describe the protocol for gen-erating the unmarked deletion of phzA1-G1: The 5_{′ region} (~1 kb in length) of the sequence flanking phzA1 was ampli-fied using the primer pair #1 and the 3′ region (~1 kb in length) of the sequence flanking phzG1 with primer pair #2 (Table 2). These flanking DNA fragments were joined using overlap extension PCR. The resulting PCR product, con-taining a deletion of phzA1-G1, was cloned into a unique SpeI site in the mobilizable plasmid pSMV10. pSMV10 is a suicide plasmid for PA14 and contains an oriR6K origin of replication that does not function in PA14 but replicates in E. coli strains containing the pir gene; a gentamicin-resistance gene (aacC1); an oriT from RP4 that allows for mobilization by E. coli strains carrying RP4-derivatives on their chromosome (E. coli WM3064); and the

counterselect-able sacB gene. The resulting deletion plasmid was trans-formed into E. coli WM3064 and mobilized into PA14 using biparental conjugation (Whiteley et al., 1999). PA14 single recombinants (merodiploid containing the intact phzA1-G1 operon and the deleted operon) were selected on LB agar containing gentamicin. Potential phzA1-G1 deletion mutants were generated by selecting for a resolved merodiploid (double recombinant) by identifying strains that grew in the presence of 10% sucrose (these strains lost the sacB-containing plasmid because sacB is toxic in the presence of sucrose). Strains with properties of a double recombination were further analysed by PCR to determine if phzA1-G1 has been deleted and one was selected. The deletion of phzA2-G2 and soxR were made the same way, using primer pairs # 3 and 4, and #5 and 6 respectively (Table 3).

Quantification of phenazine derivatives

For cultures grown in MOPS medium, PYO and PCA con-centrations were measured via HPLC analysis of culture fil-trates (0.2mm pore). Two hundred microlitres of filtrates from MOPS minimal medium cultures (aerobic cultures: 100 ml in a 500 ml Erlenmeyer flask) was loaded directly onto a Waters Symmetry C18 reverse-phase column (5mm particle size; 4.6¥ 250 mm) in a Beckman SystemGold set up with a photodiode array detector. Phenazines were separated in a gradient of water-0.01% TFA (solvent A) to acetonitrile-0.01% TFA (solvent B) at a flow rate of 1 ml min-1_{in the following method: linear gradient from 0 to}

15% solvent B from 0 to 2 min, linear gradient to 83% solvent B from 2 to 22 min, then a linear gradient to 0% solvent B from 22 to 24 min. The total method time was 25 min. Retention times for PYO and PCA averaged 10.933 and 19.918 respectively. Solutions of purified PYO and PCA were prepared at known concentrations based on their absorbance spectra at neutral pH [the extinction coefficient e of PYO at 690 nm is 4130 M-1_cm-1_{(Reszka et al., 2004);}

e of PCA at 368 nm is 11 340 M-1

cm-1

(Trutko, 1989)].

Table 2. Bacterial strains and plasmids used in this study.

Strains/plasmids Characteristics Source or reference

P. aeruginosa

PA14 Clinical isolate UCBPP-PA14 Rahme et al. (1995)

PA14Dphz1/2 PA14 with deletions of operons phzA1-G1 and phzA2-G2 This study

PA14DsoxR PA14 with a deletion in soxR This study

PA14DpqsH PA14 with an insertion in pqsH; GmR _{Mashburn and Whiteley (2005)} PA14 phzM PA14 mutant with a GmR_{cassette inserted into PA4209 (phzM) Gift from P. Cornelis, Brussels, Belgium}

PAO1 Wild-type Holloway et al. (1979)

E. coli

UQ950 E. coli DH5a l(pir) host for cloning; F-D(argF-lac)169 F80

dlacZ58(DM15) glnV44(AS) rfbD1 gyrA96(NalR_{) recA1 endA1}

spoT1 thi-1 hsdR17 deoRlpir+

D. Lies, Caltech

WM3064 Donor strain for conjugation: thrB1004 pro thi rpsL hsdS lacZ DM15RP4–1360 D(araBAD)567 DdapA1341::[erm pir(wt)]

W. Metcalf, University of Illinois Plasmids

pSMV10 9.1 kb mobilizable suicide vector; oriR6K, mobRP4, sacB, KmR_GmR _{D. Lies, Caltech} pDphzA1-G1 2 kb fusion PCR fragment containingDphzA1-G1 cloned into the SpeI site of

pSMV10; used to make the strain PA14Dphz1/2

This study pDphzA2-G2 2 kb fusion PCR fragment containingDphzA2-G2 cloned into the SpeI site of

pSMV10; used to make the strain PA14Dphz1/2

This study

Km, kanamycin; Gm, gentamicin.

Aqueous solutions at approximately 0.1 mM phenazine were also loaded in 200ml volumes onto the column and subjected to the same analysis as culture filtrates. System Gold 32 Karat Software was used to calculate the area under each peak in absorbance units in the 366 nm channel. Phenazine standards at known concentrations were used to calculate conversion factors for PYO and PCA and were 8¥ 10-6_{mM/AU and 9.5¥ 10}-6_mM/AU

respectively.

For cultures grown in LB medium, cells were removed by centrifuging aliquots in spin columns (0.2mm filter pore size) at 10 000 g. The concentration of PYO in the remaining fil-trate was determined spectrophotometrically based on its absorbance at 690 nm.

Purification of PYO

Pyocyanin was prepared from late stationary phase cultures of PA14 grown in LB based on protocols established in Chang and Blackwood (1968) and Ingledew and Campbell (1969). Six hundred and fifty millilitres of cultures were grown in 2 l flasks, then split into two 500 ml centrifuge bottles and cen-trifuged for 10 min at 8000 g, at room temperature. Superna-tants from PA14 cultures were then extracted twice with 250 ml chloroform. PYO was further purified by extraction into 2¥ 250 ml 0.01 N HCl, and extracted again into 2¥ 250 ml chloroform. The organic phase was collected and evaporated to obtain a volume of approximately 5 ml and loaded onto a Waters Sunfire C18 Prep column (5 micron

particle size; 19¥ 100 mm) in 1 ml of aliquots. Samples were subjected to the same solvent system and gradient described above, but at a flow rate of 10 ml min-1_{, using the following}

method: 100% solvent A for 0–1 min, linear gradient to 70% solvent B from 1 to 10 min, then to 100% solvent B from 10 to 11 min, followed by a linear gradient to 100% solvent A from 11 to 12 min. The total method time was 15 min. Fractions containing PYO were collected using a Beckman SC100

fraction collector, and extracted into chloroform again as described above. They were then evaporated to dryness and resuspended in MOPS minimal medium to a final concentra-tion of 5 mM.

Analysis of global gene expression using P. aeruginosa Affymetrix GeneChips

Pseudomonas aeruginosa was grown aerobically in MOPS minimal medium containing 20 mM succinate to an OD600of

0.4, then 5 ml of culture was removed, mixed with 10 ml of Bacterial RNAProtect (Qiagen), and incubated for 5 min at room temperature before cells were pelleted (10 min, 5000 g). DNA-free RNA was isolated from the cell pellet using the RNAEasy Mini kit (Qiagen) including on-column DNase treatment. RNA integrity was monitored by gel elec-trophoresis of glyoxylated samples. Preparation of labelled cDNA and processing of P. aeruginosa GeneChip arrays were performed as previously described (Schuster et al., 2003). Washing, staining and scanning of GeneChips were performed by the University of Iowa DNA Core Facility using an Affymetrix fluidics station. Gene-Chip arrays were per-formed in duplicate. Data were analysed using Microarray Suite software, and only genes exhibiting regulation levels of twofold or greater are reported. The GeneChip P. aerugi-nosa Genome Array (Affymetrix) contains probe sets for more than 5500 genes from P. aeruginosa PAO1, 199 probe sets corresponding to 100 intergenic sequences, and 117 additional genes from P. aeruginosa strains other than PAO1. Due to the strong similarity between strains PAO1 and PA14 the Array has been used for gene expression analysis of both strains (Whiteley et al., 2001; Mashburn et al., 2005). Verification of GeneChip data was performed with Q-RT-PCR. PA4205 (mexG), PA4206 (mexH) and PA2274 expression levels were examined using PA1802 (clpX) and PA3617 (recA) as constitutively expressed con-trols (primer sets given in Table 2).

Table 3. Primers used in this study.

5′ primer 3′ primer

A. Primers used for pSMV10 constructs

pSMV10-DphzA1-G1 # 1 GGactagtAGAACAGCACCATGTC cccatccactaaatttaaataTGTACC # 2 tatttaaatttagtggatgggCACCGCTACCTGCAAC GCGactagtGGGT32TCTTCGATCACTAC pSMV10-DphzA2-G2 # 3 GCGactagtGCTGATCTGGAATGGCG cccatccactaaatttaaataCAACCGTTGGTACTCTCG # 4 tatttaaatttagtggatgggCACCGCTACCTGCAAC GCGactagtGGGTTTCTTCGATCACTAC pSMV10-DsoxR # 5 gcgACTAGTgatgcgcaggaattcttc cccatccactaaatttaaataGGCCGCGAGCACGAC # 6 tatttaaatttagtggatgggGGATGCGCAGGAATTCTTC gcgACTAGTgcctcggtcacatgggc B. Primers used for Q-RT-PCR

Target gene

PA1802 (clpX) CCT GTG CAA TGA CAT CC AGG ATG GTG CGG ATC TCT TT

PA2274 GAA CGC TGC TTC AGG AAC TG GTG AAC GCC AGC AGG TGT AG

PA2275 GAC CAG GTG ATC CTT TCC AC GTA GGG ATT GAG GTC GTG CT

PA2587 (pqsH) CGG ATC GAG TTC ATC AGG A CGA ACG AGG GTA TTC CTC AG PA3617 (recA) CTG CCT GGT CAT CTT CAT CA ACC GAG GCG AGA ACT TCA G

PA3920 AGC GTC TAC CTC TGG CTG AC AGG TAC TTG CCG AGG AGG AT

PA4205 (mexG) AAC TCG CTC GAA AGC AAC TG GCT GGC CTG ATA GTC GAA CA

PA4206 (mexH) CAC CTC GGC CAG TAC CTC GAC TGG TGC TTT CGT CCA G

All primers are written in 5_{′ to 3 ′ direction.}

Quantitative reverse transcriptase PCR

Pseudomonas aeruginosa was grown in LB medium or MOPS minimal medium containing 20 mM succinate to an OD600 of 0.4 (exponential phase) or 1.4–1.8 (stationary

phase). One volume of culture was mixed with two volumes of Bacterial RNAProtect (Qiagen), incubated for 5 min at room temperature, and centrifuged for 10 min at 5000 g. Total RNA was extracted from the cell pellet using an RNAEasy Mini kit (Qiagen), according to the manufacturer’s instruc-tions, including the optional DNase treatment step. cDNA was generated using the extracted RNA as template for a Taqman (ABI Biosciences) random-primed reverse transcriptase reaction following the manufacturer’s protocol. The cDNA was used as template for quantitative PCR (Real Time 7300 PCR Machine, Applied Biosystems) using the Sybr Green detection system (Applied Biosystems). Samples were assayed in triplicate. Signal was standardized to recA using the following equation: Relative expression=2(CTstandard−CTsample)_,

where CT (cycle time) was determined automatically by the Real Time 7300 PCR software (Applied Biosystems). Primers (Integrated DNA Technologies) for Q-RT-PCR were designed using Primer3 software (Rozen and Skaletsky, 2000). Criteria for primer design were a melting temperature of 60°C, primer length of 20 nucleotides, and an amplified PCR fragment of 100 base pairs.

Acknowledgements

We are grateful to Pierre Cornelis for sending us a phzM mutant of PA14. We thank members of the Newman lab and Tracy J. LaGrassa for helpful discussions. This work was supported by grants from the Luce Foundation, Packard Foundation, and Howard Hughes Medical Institute to D.K.N. L.E.P.D. is supported by a postdoctoral EMBO Long Term Fellowship, and A.M.P-W. by an N.I.H. training grant.

References

Aendekerk, S., Ghysels, B., Cornelis, P., and Baysse, C. (2002) Characterization of a new efflux pump, MexGHI-OpmD, from Pseudomonas aeruginosa that confers resis-tance to vanadium. Microbiology 148: 2371–2381. Aendekerk, S., Diggle, S.P., Song, Z., Hoiby, N., Cornelis, P.,

Williams, P., and Camara, M. (2005) The MexGHI-OpmD multidrug efflux pump controls growth, antibiotic suscepti-bility and virulence in Pseudomonas aeruginosa via 4-quinolone-dependent cell-to-cell communication. Micro-biology 151: 1113–1125.

Budzikiewicz, H. (1993) Secondary metabolites from fluores-cent pseudomonads. FEMS Microbiol Rev 10: 209–228. Camilli, A., and Bassler, B.L. (2006) Bacterial small-molecule

signaling pathways. Science 311: 1113–1116.

Chang, P.C., and Blackwood, A.C. (1968) Simultaneous bio-synthesis of pyocyanine, phenazine-1-carboxylic acid, and oxychloroaphine from labelled substrates by Pseudomo-nas aeruginosa Mac 436. Can J Biochem 46: 925–929. Cox, C.D. (1986) Role of pyocyanin in the acquisition of iron

from transferrin. Infect Immun 52: 263–270.

Deziel, E., Lepine, F., Milot, S., He, J., Mindrinos, M.N., Tompkins, R.G., and Rahme, L.G. (2004) Analysis of

Pseudomonas aeruginosa 4-hydroxy-2-alkylquinolines

(HAQs) reveals a role for 4-hydroxy-2-heptylquinoline in cell-to-cell communication. Proc Natl Acad Sci USA 101: 1339–1344.

Deziel, E., Gopalan, S., Tampakaki, A.P., Lepine, F., Pad-field, K.E., Saucier, M., et al. (2005) The contribution of MvfR to Pseudomonas aeruginosa pathogenesis and quorum sensing circuitry regulation: multiple quorum sensing-regulated genes are modulated without affecting lasRI, rhlRI or the production of N-acyl- 1-homoserine lactones. Mol Microbiol 55: 998–1014.

Diggle, S.P., Winzer, K., Chhabra, S.R., Worrall, K.E., Camara, M., and Williams, P. (2003) The Pseudomonas aeruginosa quinolone signal molecule overcomes the cell density-dependency of the quorum sensing hierarchy, regulates rhl-dependent genes at the onset of stationary phase and can be produced in the absence of LasR. Mol Microbiol 50: 29–43.

Ding, H., Hidalgo, E., and Demple, B. (1996) The redox state of the [2Fe-2S] clusters in SoxR protein regulates its activity as a transcription factor. J Biol Chem 271: 33173–33175. Eberhard, A., Burlingame, A.L., Eberhard, C., Kenyon, G.L.,

Nealson, K.H., and Oppenheimer, N.J. (1981) Structural identification of autoinducer of Photobacterium fischeri luciferase. Biochemistry 20: 2444–2449.

Engebrecht, J., Nealson, K., and Silverman, M. (1983) Bac-terial bioluminescence: isolation and genetic analysis of functions from Vibrio fischeri. Cell 32: 773–781.

Fuqua, W.C., Winans, S.C., and Greenberg, E.P. (1994) Quorum sensing in bacteria: the LuxR-LuxI family of cell density-responsive transcriptional regulators. J Bacteriol 176: 269–275.

Gallagher, L.A., McKnight, S.L., Kuznetsova, M.S., Pesci, E.C., and Manoil, C. (2002) Functions required for extra-cellular quinolone signaling by Pseudomonas aeruginosa. J Bacteriol 184: 6472–6480.

Gaudu, P., and Weiss, B. (1996) SoxR, a [2Fe-2S] transcrip-tion factor, is active only in its oxidized form. Proc Natl Acad Sci U S A 93: 10094–10098.

Gaudu, P., Moon, N., and Weiss, B. (1997) Regulation of the soxRS oxidative stress regulon. Reversible oxidation of the Fe-S centers of SoxR in vivo. J Biol Chem 272: 5082– 5086.

Grasemann, H., Ioannidis, I., Tomkiewicz, R.P., de Groot, H., Rubin, B.K., and Ratjen, F. (1998) Nitric oxide metabolites in cystic fibrosis lung disease. Arch Dis Child 78: 49–53. Hassett, D.J., Charniga, L., Bean, K., Ohman, D.E., and

Cohen, M.S. (1992) Response of Pseudomonas aerugi-nosa to pyocyanin: mechanisms of resistance, antioxidant defenses, and demonstration of a manganese-cofactored superoxide dismutase. Infect Immun 60: 328–336. Hays, E.E., Wells, I.C., Katzman, P.A., Cain, C.K., Jacobs,

F.A., Thayer, S.A., et al. (1945) Antibiotic substances pro-duced by Pseudomonas aeruginosa. J Biol Chem 159: 725–750.

Hernandez, M.E., and Newman, D.K. (2001) Extracellular electron transfer. Cell Mol Life Sci 58: 1562–1571. Hernandez, M.E., Kappler, A., and Newman, D.K. (2004)

Phenazines and other redox-active antibiotics promote

microbial mineral reduction. Appl Environ Microbiol 70: 921–928.

Hidalgo, E., Ding, H., and Demple, B. (1997) Redox signal transduction via iron-sulfur clusters in the SoxR transcrip-tion activator. Trends Biochem Sci 22: 207–210.

Holloway, B.W., Krishnapillai, V., and Morgan, A.F. (1979) Chromosomal genetics of Pseudomonas. Microbiol Rev 43: 73–102.

Ingledew, W.M., and Campbell, J.J. (1969) A new resuspen-sion medium for pyocyanine production. Can J Microbiol 15: 595–598.

Kobayashi, K., and Tagawa, S. (2004) Activation of SoxR-dependent transcription in Pseudomonas aeruginosa. J Biochem (Tokyo) 136: 607–615.

Lazdunski, A.M., Ventre, I., and Sturgis, J.N. (2004) Regula-tory circuits and communication in Gram-negative bacteria. Nat Rev Microbiol 2: 581–592.

Lepine, F., Deziel, E., Milot, S., and Rahme, L.G. (2003) A stable isotope dilution assay for the quantification of the Pseudomonas quinolone signal in Pseudomonas aerugi-nosa cultures. Biochim Biophys Acta 1622: 36–41. McGrath, S., Wade, D.S., and Pesci, E.C. (2004) Dueling

quorum sensing systems in Pseudomonas aeruginosa control the production of the Pseudomonas quinolone signal (PQS). FEMS Microbiol Lett 230: 27–34.

Mahajan-Miklos, S., Tan, M.W., Rahme, L.G., and Ausubel, F.M. (1999) Molecular mechanisms of bacterial virul-ence elucidated using a Pseudomonas aeruginosa-Caenorhabditis elegans pathogenesis model. Cell 96: 47–56.

Mashburn, L.M., and Whiteley, M. (2005) Membrane vesicles traffic signals and facilitate group activities in a prokaryote. Nature 437: 860–864.

Mashburn, L.M., Jett, A.M., Akins, D.R., and Whiteley, M. (2005) Staphylococcus aureus serves as an iron source for Pseudomonas aeruginosa during in vivo coculture. J Bac-teriol 187: 554–566.

Mavrodi, D.V., Bonsall, R.F., Delaney, S.M., Soule, M.J., Phillips, G., and Thomashow, L.S. (2001) Functional analy-sis of genes for biosyntheanaly-sis of pyocyanin and phenazine-1-carboxamide from Pseudomonas aeruginosa PAO1. J Bacteriol 183: 6454–6465.

Mavrodi, D.V., Blankenfeldt, W., and Thomashow, L.S. (2006) Phenazine compounds in fluorescent Pseudomo-nas Spp. biosynthesis and regulation. Annu Rev Phyto-pathol (in press).

Mazzola, M., Cook, R.J., Thomashow, L.S., Weller, D.M., and Pierson, L.S., 3rd (1992) Contribution of phenazine antibi-otic biosynthesis to the ecological competence of fluores-cent pseudomonads in soil habitats. Appl Environ Microbiol 58: 2616–2624.

Palma, M., Zurita, J., Ferreras, J.A., Worgall, S., Larone, D.H., Shi, L., et al. (2005) Pseudomonas aeruginosa SoxR does not conform to the archetypal paradigm for SoxR-dependent regulation of the bacterial oxidative stress adaptive response. Infect Immun 73: 2958– 2966.

Palmer, K.L., Mashburn, L.M., Singh, P.K., and Whiteley, M. (2005) Cystic fibrosis sputum supports growth and cues key aspects of Pseudomonas aeruginosa physiology. J Bacteriol 187: 5267–5277.

Park, W., Pena-Llopis, S., Lee, Y., and Demple, B. (2006) Regulation of superoxide stress in Pseudomonas putida KT2440 is different from the SoxR paradigm in Escherichia coli. Biochem Biophys Res Commun 341: 51–56.

Passador, L., Cook, J.M., Gambello, M.J., Rust, L., and Iglewski, B.H. (1993) Expression of Pseudomonas aerugi-nosa virulence genes requires cell-to-cell communication. Science 260: 1127–1130.

Pearson, J.P., Passador, L., Iglewski, B.H., and Greenberg, E.P. (1995) A second N-acylhomoserine lactone signal produced by Pseudomonas aeruginosa. Proc Natl Acad Sci USA 92: 1490–1494.

Pesci, E.C., Milbank, J.B., Pearson, J.P., McKnight, S., Kende, A.S., Greenberg, E.P., and Iglewski, B.H. (1999) Quinolone signaling in the cell-to-cell communication system of Pseudomonas aeruginosa. Proc Natl Acad Sci USA 96: 11229–11234.

Pomposiello, P.J., Bennik, M.H., and Demple, B. (2001) Genome-wide transcriptional profiling of the Escherichia coli responses to superoxide stress and sodium salicylate. J Bacteriol 183: 3890–3902.

Price-Whelan, A., Dietrich, L.E., and Newman, D.K. (2006) Rethinking ‘secondary’ metabolism: physiological roles for phenazine antibiotics. Nat Chem Biol 2: 71– 78.

Rahme, L.G., Stevens, E.J., Wolfort, S.F., Shao, J., Tomp-kins, R.G., and Ausubel, F.M. (1995) Common virulence factors for bacterial pathogenicity in plants and animals. Science 268: 1899–1902.

Reszka, K.J., O’Malley, Y., McCormick, M.L., Denning, G.M., and Britigan, B.E. (2004) Oxidation of pyocyanin, a cyto-toxic product from Pseudomonas aeruginosa, by microp-eroxidase 11 and hydrogen peroxide. Free Radic Biol Med 36: 1448–1459.

Rozen, S., and Skaletsky, H. (2000) Primer3 on the WWW for general users and for biologist programmers. Methods Mol Biol 132: 365–386.

Schuster, M., Lostroh, C.P., Ogi, T., and Greenberg, E.P. (2003) Identification, timing, and signal specificity of Pseudomonas aeruginosa quorum-controlled genes: a transcriptome analysis. J Bacteriol 185: 2066–2079. Sciara, G., Kendrew, S.G., Miele, A.E., Marsh, N.G., Federici,

L., Malatesta, F., et al. (2003) The structure of ActVA-Orf6, a novel type of monooxygenase involved in actinorhodin biosynthesis. EMBO J 22: 205–215.

Stover, C.K., Pham, X.Q., Erwin, A.L., Mizoguchi, S.D., War-rener, P., Hickey, M.J., et al. (2000) Complete genome sequence of Pseudomonas aeruginosa PA01, an opportu-nistic pathogen. Nature 406: 959–964.

Touati, D. (2000) Sensing and protecting against superoxide stress in Escherichia coli – how many ways are there to trigger soxRS response? Redox Rep 5: 287–293. Trutko, S.M. (1989) The physiological role of phenazine

pig-ments synthesized by the bacteria Pseudomonas aureofaciens. Biochemistry (Mosc) 54: 1092–1098. Warren, J.B., Loi, R., Rendell, N.B., and Taylor, G.W. (1990)

Nitric oxide is inactivated by the bacterial pigment pyocyanin. Biochem J 266: 921–923.

Wells, I.C. (1952) Antibiotic substances produced by Pseudomonas aeruginosa; syntheses of Pyo Ib, Pyo Ic,

and Pyo III. J Biol Chem 196: 331–340.

Whiteley, M., Lee, K.M., and Greenberg, E.P. (1999) Identi-fication of genes controlled by quorum sensing in Pseudomonas aeruginosa. Proc Natl Acad Sci USA 96: 13904–13909.

Whiteley, M., Bangera, M.G., Bumgarner, R.E., Parsek, M.R., Teitzel, G.M., Lory, S., and Greenberg, E.P. (2001) Gene expression in Pseudomonas aeruginosa biofilms. Nature 413: 860–864.

Supplementary material

The following supplementary material is available for this article online:

Figure S1. Characterization of PA14 phzM

Figure S2. Deletion of soxR does not affect planktonic PA14 growth.

This material is available as part of the online article from http://www.blackwell-synergy.com